A CDC paper by Tom Shimabukuro in the New England Journal of Medicine, when you do the math, made it look like the first-trimester miscarriage rate could be as high as 84%, though the true figure could not really be told.

Nov 7, 2026

Previously: A CDC paper by Tom Shimabukuro in the New England Journal of Medicine, when you do the math, made it look like the first-trimester miscarriage rate could be as high as 84%, though the true figure could not really be told. [Distorted]: The NEJM paper itself never suggested a figure like 84%. Critics computed 82 to 91% by dividing its 104 miscarriages by only the roughly 127 early pregnancies with known outcomes, ignoring that most early pregnancies had not finished yet, and Full Fact reports that article was withdrawn. The NEJM authors did correct their own denominator, but follow-up of the remaining pregnancies found a miscarriage rate of about 13%, in line with the general population. This verdict would be wrong if the original NEJM paper or its correction itself stated a first-trimester miscarriage rate near 84%.

Now: A CDC paper by Tom Shimabukuro in the New England Journal of Medicine, when you do the math, made it look like the first-trimester miscarriage rate could be as high as 84%, though the true figure could not really be told. [Distorted]: McCullough hedges that incomplete data only looked like an 84% risk. That is not a valid risk estimate: early ongoing pregnancies cannot be left out of the denominator just because they have not finished. Subsequent registry follow-up estimated 14.1% loss at 6-19 weeks, 12.8% after age standardization, within expected background ranges. This observational evidence does not establish zero risk or measure losses before six weeks.

Replace the secondary fact check with the registry analysis and preserve the speaker's incomplete-follow-up qualification.